%latex.default(object = resultC, title = "", file = "tables/experimentC_learning_M3.tex",     label = "tb_experimentC_learning_M3", caption = captionC_M3,     insert.bottom = noteC_M3, first.hline.double = FALSE, rowname = NULL,     colheads = clabelsPeriod, col.just = c("l", rep("c", ncol(resultC) -         1)), n.rgroup = c(1, 8, 2), longtable = FALSE, center = "centering")%
\begin{table}[!tbp]
\caption{(Lab Experiment) Learning Effect: Kicking out the Least Preferred Leading Party ($M = 3$)\label{tb_experimentC_learning_M3}} 
{\centering
\begin{tabular}{lcccccccc}
\hline
\multicolumn{1}{c}{}&\multicolumn{1}{c}{}&\multicolumn{1}{c}{Period 1--6}&\multicolumn{1}{c}{}&\multicolumn{1}{c}{Period 4--9}&\multicolumn{1}{c}{}&\multicolumn{1}{c}{Period 7--12}&\multicolumn{1}{c}{}&\multicolumn{1}{c}{All Periods}\tabularnewline
\hline
&&&&&&&&\tabularnewline
District magnitude ($M$)&$$&3&$$&3&$$&3&$$&3\tabularnewline
\hline
&&&&&&&&\tabularnewline
\multirow{2}{*}{Pref. for the runner-up party}&$$&2&$$&2.588&$$&2.455&$$&2.188\tabularnewline
&$$&(0.165)&$$&(0.225)&$$&(0.226)&$$&(0.132)\tabularnewline
Pref. for the most preferred&$$&-1.632&$$&-1.495&$$&-1.935&$$&-1.783\tabularnewline
~~leading party&$$&(0.256)&$$&(0.321)&$$&(0.335)&$$&(0.201)\tabularnewline
Pref. for the second-most preferred&$$&\textbf{0.027}&$$&\textbf{-0.323}&$$&\textbf{-0.682}&$$&\textbf{-0.376}\tabularnewline
~~leading party&$$&\textbf{(0.455)}&$$&\textbf{(0.451)}&$$&\textbf{(0.452)}&$$&\textbf{(0.314)}\tabularnewline
Pref. for the third-most preferred&$$&\textbf{-0.146}&$$&\textbf{-0.085}&$$&\textbf{0.210}&$$&\textbf{0.149}\tabularnewline
~~leading party&$$&\textbf{(0.418)}&$$&\textbf{(0.405)}&$$&\textbf{(0.399)}&$$&\textbf{(0.285)}\tabularnewline
\hline
&&&&&&&&\tabularnewline
Number of observations&$$&1242&$$&1108&$$&1028&$$&2270\tabularnewline
Number of clusters&$$&392&$$&391&$$&407&$$&799\tabularnewline
\hline
\end{tabular}}
\parbox{0.98\textwidth}
{Notes: Coefficients and standard errors are multiplied by 100. 
The outcome variable is voting for the runner-up
against one of the leading parties.
The number of viable parties is included as covariates.
Cluster-robust standard errors in parentheses.}\end{table}
